Remi van Trijp gives LingCircle talk on Construction Grammar
The Department of Linguistics is pleased to announce the following talk by, who is visiting us from Sony Computer Science Laboratories inParis.
"What Are Constructions and What Can They Do?"
Wednesday, February 20
4:00-5:30pm
Clare Small 209
Abstract:
Construction grammar grew out of the need to model the whole of languageinstead of distinguishing core linguistic expressions from peripheral ones (Fillmoreet al., 1988; Kay and Fillmore, 1999), and has since then established itselfas the grammatical embodiment of cognitive-functional linguistics (Croft andCruse, 2004). Its central claim that all linguistic knowledge can be representedas form-meaning mappings – called constructions – has been embraced in bothdata-oriented and experiment-driven subdisciplines such as language acquisition(Dabrowska et al., 2009; Diessel, 2004; Tomasello, 2003), corpus linguistics (Hilpert,2015; Stefanowitsch and Gries, 2003; Zeschel, 2012), historical linguistics (Bardalet al., 2015; Colleman, 2016; Coussé et al., 2018; Fried, 2009; Van de Velde et al.,2013; Van Goethem, 2017), sociolinguistics (Höder, 2014; Hollmann and Siewierska,2007), psycho- and neurolinguistics (Barrès, 2017; Dominey et al., 2006; Perekand Goldberg, 2017), computational and formal linguistics (Bergen and Chang,2005; Boas and Sag, 2012; Michaelis, 2004; Steels, 2011) and artificial intelligence(Beuls and Steels, 2013; Steels, 2004; Van Eecke and Beuls, 2017).
As is often the case, however, it takes time before the potential of an innovationis fully explored and understood. Early movies, for example, strongly mimickedtheater and used long and static shots before film makers developed their owncinematic “grammar”. A similar process happens in science, and while constructiongrammar is already too mature to be directly compared to early cinema, the formaland computational properties of its most important data structure are not yetcompletely worked out. As a result, construction grammar has become an umbrellaterm for all linguistic studies that roughly agree on what Bill Croft (2005) dubbedvanilla construction grammar, but more precision is needed in order to preventa babelesque confusion from installing itself in the field and thereby impedingmuch-needed breakthroughs.
In this presentation, I will try to offer a more precise perspective on what constructionsare and what they can do. More specifically, I will look at the representationaland algorithmic properties of constructions. The goal of the presentation is therefore not to favor one or the other analysis, but simply to elicit more clarityabout which analyses are possible and which criticisms on constructional analysesare valid concerns and which are not. In order to substantiate my claims, allanalyses are accompanied by a concrete computational implementation in FluidConstruction Grammar (FCG; Steels, 2011), an open-source computational platformfor exploring issues in constructional language processing and learning.